A multitude of problems exist in biomedical science that are too complex to be solved by experimentation or mainframe power. Among these, the structure of biologically important macromolecules is a preeminent one. Obvious examples are the interferons, the products of all but one proto- oncogene, and all receptors for hormones and other biological messengers. Computation modeling fills the gap left by experimental methods and yields insights into mechanisms and structures. The computational resources described in this proposal are uniquely suited to examine the biological effects of these molecules. More specifically, the aims of the Cornell Theory Center NIH Resource are: o To build on the strengths of the Theory Center to promote collaborative research endeavors among biologists, chemists, physicists, computer scientists, and numerical mathematicians, pushing algorithm developments and software advances that will lay the basis for further advances in biomedical science; o To provide to biomedical researchers access to highly parallel computing environments that have the potential to achieve very high levels of computing power at new levels of flexibility and cost/performance and which are easily duplicated in researchers' laboratories; o To achieve scientific advances in protein folding and the study of other biologically active macromolecules areas that would be impossible without the computing capability of the Resource; o To develop methodologies and assess existing methodologies for writing programs to make portability between parallel systems possible, so that newer and more powerful machines can immediately be put to use; o To broadly disseminate the results both of research conducted on the Resource and technological developments within it; o To use the knowledge gained from developing biomedical molecular structure applications for highly parallel systems to further enhance parallel technology.